Evaluation of the Information Flow Control in the Fog Computing Model

Abstract

In the IoT (Internet of Things), data are exchanged among subjects and objects in devices through manipulating objects. Even if subjects manipulate objects in accordance with the CBAC (Capability-Based Access Control) model, the subjects can get data which are not allowed to be gotten by the subjects, i.e. illegal information flow and late information flow occur. Hence, the OI (Operation Interruption) and TBOI (Time-Based OI) protocols where operations occurring illegal and late types of information flows are interrupted are implemented. Moreover, capability token selection algorithms are proposed and applied to the protocols. The protocols are implemented and evaluated in terms of the request processing time, communication traffic, and electric energy consumption. However, the more number of operations are interrupted to prevent both types of illegal and late information flows because the amount of data kept by entities monotonically increases through manipulating objects in the protocols. Therefore, reduction of the number of operations interrupted is important. For this aim, an FC (Fog Computing) model of the IoT where data from devices are processed in a fog layer and the processed data are sent to subjects is considered in this paper. In the evaluation, it is shown that the number of operations interrupted is reduced in the FC-based protocols compared with the conventional protocols.